About CA
Contact CA

Search Journal

Editorial Board

Permission to Reprint


Site Map


Replicating Paintings

  Matteo Ravasio

In this paper, I discuss cases of replication in the visual arts, with particular focus on paintings. In the first part, I focus on painted copies, that is, manual reproductions of paintings created by artists. Painted copies are sometimes used for the purpose of aesthetic education on the original. I explore the relation between the creation of painted copies and their use as aesthetic surrogates of the original artwork and draw a positive conclusion on the aesthetic benefits of replica production by artists. A skeptical conclusion follows regarding the use of such replicas as surrogates for the original painting. The second part of the paper concerns mechanically produced replicas, such as photographs and 3-D prints. On the basis of some of the claims made in the first part, I set conditions that mechanically produced replicas need to meet in order to function as aesthetic surrogates of an original. I argue that perfect aesthetic surrogates are either already available or at least possible. I conclude by considering two possible objections.

Key Words
aesthetic education; copies; forgeries; painting; replication


1. Introduction

The exhibition, Replicating Genius: Impressionism 1874, curated by Nathaniel Dunn, was held at the Gus Fisher Gallery in Auckland, New Zealand, from November 11-16, 2016. It consisted entirely of replicas of paintings, from the first impressionist exhibition of 1874, commissioned by the curator from professional artists.[1] In the curatorial notes, Dunn writes: “Many people, myself included, do not have the means to travel the world to look at every artwork they would like. …So how can great art be brought to a more local environment?”[2] His answer to this question is that replicas may provide a satisfactory substitute for the appreciation of original artworks. Unlike printed reproductions in a book, replicas can be the exact size of the original, or close enough to give the viewer a better impression of the original’s size. With regard to the purpose of Replicating Genius, Dunn states, “The ultimate goal is to broaden art education and appreciation.”[3] It is evident from this statement of intent that Dunn believes replicas provide some aesthetic insight into the qualities of the original painting.[4] He is not alone in holding this belief, as he observes, “…replication has a long history in fine arts …. Copies and copying were used as didactic tools, and that is how they are being used today.”[5]

Replicating Genius features two relevant aspects of replication: the creation of painted replicas by artists and their use to replace an original. The first part of this paper explores the relation between these two aspects and draws a positive conclusion on the aesthetic benefits of replica production by artists, an aspect largely ignored by the extant literature. A skeptical conclusion follows regarding the use of such replicas as a surrogate for the original painting in order to promote aesthetic education. The second part of the paper concerns mechanically produced replicas, such as photographs and 3-D prints. On the basis of the observations developed in the first part, I set some conditions that mechanically produced replicas need to meet in order to function as aesthetic surrogates of an original. I argue that perfect aesthetic surrogates are either already available or at least possible. I conclude by considering two possible objections and concede that, at least under one possible interpretation, perfect copies may lack one aspect of the aesthetic experience of the original, although this would not interfere with their use in aesthetic education.

2. Appreciation of replicas

The main reason to produce replicas is that they allow the viewer to experience some of the aesthetic properties of the original. This goal lies behind exhibitions such as Replicating Genius, as illustrated by the curatorial statement quoted above. A similar function is performed by the most common example of replication, that is, photographs of artworks contained in any art history book and exhibition catalogues. In other words, replicas function as aesthetic surrogates of the original. This function is what allows them to perform their role in aesthetic education.[6]

Different strategies may be employed to replace the experience of originals. The exhibition, Replicating Genius, chose the way of painted copies. Professional artists sat in front of a high-quality photograph of the original and reproduced it as closely as possible on canvases of the same or similar size as the original. These artists often carried out extensive research to become more familiar with the painting methods of the artist they set out to replicate. However, other strategies are available. In 2006, Factum Arte, on commission from the Giorgio Cini Foundation, used state-of-the-art 3-D scanners and printers to produce a copy of Paolo Veronese’s The Wedding at Cana (1563).[7] The painting has been absent from its original Venetian location, the Basilica di San Giorgio Maggiore, since Napoleonic troops looted it in 1797. The copy was installed in the Basilica’s refectory, the exact place intended for the original, now in the Louvre.

It is not my aim to assess which of the available replication strategies is the most effective. A definitive answer to this question is partly tied to the answer to the questions of how the surrogate experience is produced and in what relation the copy stands to the original.[8] However, the success of the Veronese replica commands attention for the possibility of widespread use of recent technological developments in order to produce copies that are barely distinguishable from the original, even upon close scrutiny. I return to this possibility at the end of the paper and compare it with the use of painted copies.

3. Replica production: painted copies

This section is devoted to the neglected point of view of the copyist, so far an unexplored topic in philosophical literature. For the sake of simplicity, I will talk about copying paintings. The paradigm here is that of the traditional copying process in which an artist sits in front of a painting, or a suitably accurate photograph of it, and replicates it in the same medium, or the closest one available. My claims will concern the relation of this process to aesthetic education.

In discussing the harm made by forgeries, Sherri Irvin describes aesthetic understanding as being essentially a bootstrapping process, constrained by the limits of our background knowledge and perceptual discrimination.[9] She claims that, in the absence of objective, axiom-like principles on which to ground aesthetic discrimination and evaluation, we are left with a framework in which our perceptual abilities are informed by relevant knowledge concerning the artwork in question. Such informed perceptual acquaintance with artworks will, in turn, refine our capacity to situate works in their proper historical category, and this will allow us to experience new artworks with renewed historical and contextual understanding.

On the basis of this, Irvin draws two main consequences regarding forgeries. First, forgeries corrupt aesthetic understanding, as they invite us to search for aesthetic value in the wrong places, either because they possess none or because the incorrect attribution sets up a network of expectations connected with an artist’s known work. Insofar as they do so, forgeries meddle with the bootstrapping process that is at the heart of aesthetic understanding. Second, known forgeries, that is, forged works that have been recognized as such, are capable of honing our aesthetic understanding, as their inauthentic character helps critics to focus on the differences between the forged work and the original. Discussing Van Meegeren’s famous Vermeer fakes, Irvin claims that:

Art historians and others have gradually been able, by looking carefully at the forgeries in relation to the original works, to recognize the ways in which aesthetic understanding was distorted before the forgery was discovered, and to refine their understanding of the true characteristics of the various periods of Vermeer’s production.[10]

Thus, the fresh look on originals that is possible whenever a forger is unmasked not only restores the damage in aesthetic understanding dealt by the fake, it also positively contributes to a better understanding of the original work by providing a perceptual benchmark against which to judge it. In 1983, Hope B. Werness set the characteristics of Vermeer’s known paintings and Van Meegeren’s fakes against each other in a comparative table. Once the inauthentic nature of the latter paintings had been exposed, it was relatively easy for a critic to spot the differences between Van Meegeren’s forgeries and genuine Vermeers, in luminosity, anatomical accuracy, composition, and other crucial aspects.[11]

The upshot of Irvin’s claims for my discussion is twofold. On the one hand, it is clear that replicas share with known forgeries the beneficial effects on our aesthetic understanding, without posing the same dangers as forgeries. Replicas share with forgeries the relevant relation to the original by possessing relevant perceptual similarities, up to being perceptually indistinguishable. However, just as with known forgeries, they cannot deceive us into thinking they are an instance of the original artist’s work. Therefore, they may similarly function as a perceptual benchmark from which to judge the aesthetic and non-aesthetic properties of the original.[12] On the other hand, and more interestingly, Irvin’s characterization of aesthetic understanding can shed a light on replica production, which, as I have previously claimed, has been largely neglected in the philosophical literature on forgeries and copies. My claim is that painting copies of artworks may foster aesthetic understanding, in that it forces the copyist down a bootstrapping process such as the one described by Irvin.

Consider the following example. Suppose an artist wanted to replicate Carel Fabritius’ The Goldfinch (1654), currently at the Mauritshuis. The painting famously depicts a pet goldfinch perched on its feeder, to which the bird is chained. Generally speaking, two main characteristics of the painting stand out. On the one hand, our attention is captured by the loose and suggestive treatment of the bird’s plumage.[13] On the other hand, the feeder has a trompe l’oeil character and is rendered with remarkable attention to tonal values and shading. Armed with a basic art historical understanding of the painting’s context of production, the artist begins to paint the replica.

The overall naturalistic feel of the painting, especially evident in the trompe l’oeil effect of the feeder, may suggest to the artist that the bird should be rendered with abundance of detail. Upon closer scrutiny, however, the artist realizes that the goldfinch’s feathers are painted in rapid brushstrokes. The dark plumage of the bird’s wing is visible through the yellow feathers on top, and the black line in the middle of the yellow patch has likely been obtained by scratching the surface of the panel with the tip of the brush’s handle, a technique used by Rembrandt to paint hair curls in his self-portraits. These features of the painting show its connection with Rembrandt’s style. Awareness of such stylistic and historical properties of the work foster a more detailed observation and reproduction of such features. For instance, the artist might pay attention to the main direction of the brushstrokes in the original and then replicate a similarly free treatment in painting the copy.

Moving on to the feeder, and having already noticed its illusionistic, three-dimensional character, the artist pays attention to tonal values and shading. While rendering the shadow cast by the feeder on the cream-colored wall, the artist notices how subtly Fabritius has rendered the effect of the feeder’s local color on the overall color of the shadow. The blue-purplish feeder and the yellowish-grey wall mix into a shadow that almost looks green at some points. This understanding of reflected light and its effect on shadow endows the painting with an atmospheric subtlety that reminds the copyist of Vermeer’s Milkmaid (1657-58).

Carel Fabritius’ The Goldfinch (1654)

Remember that the bootstrapping framework described by Irvin included background knowledge and perceptual abilities. The case I have just described shows how these two elements interact and support each other in the process of replication. As the artist copies the appearance of the painting, she or he becomes more aware of the properties it instantiates and more able to determine the artwork’s position among other similar works. In turn, this awareness allows the artist to better notice and reproduce the non-aesthetic features of the work that are responsible for the aesthetic properties. However, the process of copying the painting’s appearance always starts from a prior understanding of the aesthetically relevant elements of the picture that are related to the work’s classification as, say, figurative seventeenth-century oil painting.

4. Painted copies and aesthetic understanding

I have claimed that painted copies instantiate the bootstrapping process of aesthetic understanding described by Irvin. I have also claimed that replicas, just like known forgeries, have the potential to refine our aesthetic understanding. From this standpoint, it is worth going back to the issue regarding the use of painted copies as a surrogate of the original painting. In fact, a problem ensues when painted replicas are used for the purpose of aesthetic education on the original painting. Recall that this is what the exhibition Replicating Genius set out to do. According to its curator, painted copies can be a successful means to put a vast number of people in direct contact with masterpieces scattered around the globe. In this case, replicas are used, in lieu of the original, for the purpose of aesthetic education concerning that original work. The problem can be described as follows.

In order for the bootstrapping process to work, the relevant art-historical concepts must be brought to bear on the right sort of perceptual content. Take the case in which I examine Fabritius’ Goldfinch under the category Dutch painting. Careful scrutiny of the perceptual surface may reveal finer-grained perceptual qualities than the ones available to the untrained eye. For instance, the loose rendering of the plumage will stand out as one of the picture’s highlights. The problem for painted copies is that it is impossible to preemptively determine what would count as the relevant perceptual content. Potentially any perceivable feature of the picture may be subject to finer-grained perceptual discrimination, once the concepts brought to bear on the picture are refined enough and our perceptual abilities have been correspondingly sharpened.

Now, the perceptual experience of painted copies is non-trivially different from the perceptual experience of originals. Therefore, uncontrolled exposure of an untutored audience to painted replicas may have long-term effects on aesthetic understanding as harmful as those of undetected forgeries. Painted replicas, when they are the sole source of acquaintance with the original painting, face the viewer with a perceptual surface that, in virtue of its being non-trivially different from that of the original work, is unfit to become part of the bootstrapping process of aesthetic understanding described by Irvin. Thus, while painted copies may be usefully compared with the original, or with a perceptually indistinguishable reproduction, a museum of painted copies would pose a threat to aesthetic education similar to that of a Vermeer exhibition constituted entirely by Van Meegerens.[14]

According to the curator of Replicating Genius, an additional motivation for exhibiting painted copies is that they can bring to the public works that are hard to see in person because they are owned by private collectors or are rarely, if ever, on display. However, if it is true that comparison with the original work is necessary in order for painted copies to foster aesthetic education, then the detrimental effect of painted copies may be compounded when the original artworks are not easily accessible. If all of the above is correct, caution should be used when painted replicas are employed for the purpose of aesthetic education and, especially in the light of recent technological developments, mechanically produced copies may be preferable.

5. Replica production: mechanically produced replicas

The skeptical conclusion I reached concerning the educational value of painted copies suggests focusing on mechanically produced replicas. The prototypical case here is the photograph. Art history books and exhibition catalogues present us with the aesthetic properties of the original by reproducing photographs of them. But how do photographs allow us to experience the properties of the original? Can they always do so? Or do they have limitations?

Robert Hopkins suggests a promising hypothesis, without endorsing it. Photographs function as an aesthetic surrogate because they are transparent.[15] By looking at a reproductive photograph, we engage with the original work. Copies, Hopkins observes, share an important feature with photographs in that they do not need to possess all of the properties possessed by the object they depict. I can see that Mount Cook is majestic from a photograph of it but the photograph need not itself be majestic. Hopkins, however, has a reason to doubt that transparency can thoroughly explain how copies allow us to experience the original. I will return to his qualms in a moment. Before that, it is worth observing how the use of photographic reproductions as aesthetic surrogates does not preserve aspects of the experience of the original that are arguably aesthetically relevant.

Barbara Savedoff has described a variety of ways in which the experience of an artwork through a photograph is different from that of the original.[16] A photograph often differs in size from the original, and this means that reproductions do not occupy our visual fields in the same way as the originals do. The surface of photographic reproduction is glossy paper or a computer screen, and the materiality of paint is lost, along with the importance of its texture; stereoscopic vision is frustrated by a flat photograph that presents as two-dimensional what in fact is a three-dimensional brushstroke; colors and values of the painting are only shown in a fixed light. Finally, “We lose the ability to move closer and farther away. This prevents us from discovering the tension between a painting’s visual effect and the surface which allows that effect.”[17] All of these aspects are arguably relevant to the overall aesthetic experience of a painting yet they are often absent from our experience of transparent photographs of an original artwork.

To complicate matters, Hopkins notices a problem for transparency in photographs of artworks. Whereas in a standard photograph of a picture we would see a picture that depicts a certain scene, reproductions of artworks do not seem to offer such an experience.[18] It would seem, Hopkins claims, that we simply see the scene the painting is depicting without seeing the painting itself. If this is the case, then transparency might be the wrong relation altogether with which to explain how reproductions of artworks may function as aesthetic surrogates. Notice that, regardless of how one may be able to overcome the limitations of photographs identified by Savedoff, the difference Hopkins is pointing to would still be present. In fact, Savedoff’s concerns with our readiness to consider photographs as the artwork itself nicely fits with Hopkins’s observation. Photographs of paintings limit our appreciation to an appreciation of a photograph, as opposed to appreciation of the original painting.

Transparency may well be the way in which reproductions of the garden-variety acquaint us with some of the aesthetic properties of originals. However, transparency is not enough for photographs to function as aesthetic surrogates of an original. Regardless of this, there might still be reproduction of such kind that would allow us to bridge the gaps between the experience of the copy and that of the original. The remainder of the paper is devoted to exploring this possibility. 

6. Two gaps to bridge

I have so far claimed that the transparent character of photographs may be exploited in order to experience some of the original’s aesthetic properties. This might, after all, be the reason why photographic reproductions play an important role in art historical studies. However, transparency alone is at best of limited use if our goal is to surrogate the aesthetic experience of the picture, as it does not secure a number of important features of such experience, such as actual size, texture, and so on.

The absence of such features has two sources. On the one hand, transparent pictures of artworks are not perceptually identical to the original. Their size, texture, and the way they fill the surrounding space are not the same as the original picture. Call this the perceptual gap. On the other hand, transparent pictures are obviously different from the object they represent. At the most basic level, they differ in their causal history and spatio-temporal properties. For example, Veronese’s Nozze is in Paris, whereas its 3-D printed replica is in Venice. Because of this ontological difference, they lack properties possessed by the original that are normally considered relevant to aesthetic appreciation. While Picasso’s Demoiselles D’Avignon is a groundbreaking work, no copy of it, regardless of its accuracy, will be similarly groundbreaking. Call this difference the ontological gap.[19]

7. How to bridge the gaps

For a copy to surrogate the experience of an original, it would mean to overcome both these shortcomings. I will consider them in turn.

The perceptual gap. The problem with the perceptual gap is that copies are often non-trivially perceptually different from originals. We saw this in the case of painted copies, and we found the same problem in the case of transparent pictures. These are, in most cases, unable to provide various aspects typical of the perceptual experience of the original, such as the way the painting fills our visual field or the qualities dependent on stereoscopic vision. Moreover, the impossibility to foresee how refinements in perceptual discrimination will influence our future experience of pictures makes it hard to dismiss any perceptual difference as trivial.

The Veronese replica, however, should make us pause. Although no absolute proof has been produced that no mistakes occurred in the production of that copy, the technology employed and the care taken are such that even the most trained eye might not be able to tell apart the copy in Venice and the Paris original. More importantly, technology might soon reach a level of sophistication, if it hasn’t done so already in the Nozze case, that will allow us to produce copies with differences in size, color, texture, and 3-D shape that are small enough to be undetectable to the naked human eye. I will refer to such cases as perfect copies. A perfect copy would be only trivially perceptually different from the original, as the sort of appreciation that is normally appropriate in the case of visual arts does not require appreciation of details invisible to the human naked eye, although of course such details may be relevant when determining technical aspects, such as the chemical composition of paints. Thus, if the perceptual gap hasn’t already been bridged by the bold Veronese replication carried out by Factum Arte, it is at least conceivable that it will be bridged in the future.

The ontological gap. The problem posed by the ontological gap is that a copy, even though it may be indistinguishable from the original, is still a different object from it and lacks the right sort of causal history. For this reason, it cannot possess some of the properties that derive from the original’s significance, in the course of art-historical development, and as a member of the appreciative category it belongs to. I will examine a strong solution to this problem and reject it. Subsequently, I will propose a weaker, more defensible way to bridge the ontological gap.

Gregory Currie has argued that we should extend to the visual arts a view that is standardly accepted in the case of literature. Each accurate copy of Dickens’s Oliver Twist affords the same aesthetic experience and counts as an instance of the book. Likewise, each perceptually indistinguishable copy of a painting should count as an instance of the original work. He calls this the instance multiplicity hypothesis.[20]

Currie’s proposal has also the merit of being compatible with contextualism, that is, with the widely accepted idea that aesthetic properties depend not only on the perceptual properties of an object but also on the context in which the work was created and the category to which it belongs. Under the instance multiplicity hypothesis, perfect copies of paintings are no more harmful to contextualism than perfect copies of literary works.[21] All we need to do is to know that the instance we are examining is related to an action performed at a certain time and that the artwork belongs to a determinate appreciative category. These advantages, however, come at a price. Using the instance multiplicity hypothesis as a way to bridge the ontological gap, indeed, to eliminate it altogether, requires commitment to a suspicious ontological framework that, among other things, threatens the intuitively sound distinction between autographic and allographic art forms. For these reasons, it is worth looking for a solution that does not entail any controversial ontological commitment.

The weaker solution I wish to defend is the following. In appreciating a perfect copy of an artwork, we treat the object in front of us as if it were the original, that is, as if it had the causal history and art-historical properties that are relevant to the original artwork. In doing so, we intentionally disregard known differences between the original and the copy in front of us. For instance, the original Veronese was painted by hand, whereas the copy is 3-D printed.

Two observations are in order at this stage. First, abstracting from some of the properties of an object in order to treat it as a surrogate for another is a familiar aspect of scientific modeling. In this sense, the fictionalist proposal I sketched above is just another instance of our capacity to gain knowledge and expertise regarding an object by dealing with things other than the object itself. Needless to say, the conditions imposed by copies of artworks are more stringent than the ones typical of a scientific model. Whereas the latter may differ in many respects from the modeled object, a perfect copy needs to be perceptually indistinguishable from the original, for the reasons outlined above.

Second, and in connection with the point just made, note that I am limiting this fictionalist proposal to perfect copies, that is, copies that are perceptually indistinguishable from the original. The fictional stance towards the object’s identity, that is, the fact that we make-believe that the object in front of us has the same causal history as the original, does not interfere with the appreciative process and is therefore a perfect surrogate of the experience of the original. Here is why. Recall the bootstrapping process of aesthetic understanding where an artwork’s aesthetic properties are determined by the way in which the relevant appreciative categories are brought to bear on the non-aesthetic features of the object. Such informed scrutiny, in turn, slowly influences our understanding and application of art-historical categories. Each successive encounter with an artwork is an occasion for our perceptual discrimination abilities to be refined in the light of the refinement in the understanding of appreciative categories enabled by past appreciative experiences.

A perfect copy is, by definition, perceptually indistinguishable from the original. Hence, no possible refinement in perceptual discrimination would result in the perfect copy being incapable of functioning as part of the bootstrapping process. On the other hand, the fictional stance allows the viewer to bring to bear the relevant appreciative categories on the perceptual content. Both aspects of the bootstrapping process that allow the appreciation of the original to foster aesthetic understanding are preserved under this account. Perfect copies are therefore suitable to the purpose of aesthetic education concerning the original.

8. Two possible objections

An initial objection might be that this proposal fails to accommodate a widely accepted contextualist view of aesthetic properties. Once more, contextualism is the idea that aesthetic properties depend on non-aesthetic perceptual properties of an object plus the art-historical context in which the work was created and the category to which it belongs. For instance, the expressiveness of Titian’s reds depends as much on the specific hue and distribution of paint on his canvases as it does on its art-historical place in Venetian oil painting.

It might be thought that, as I take two perceptually indistinguishable objects to be able to provide the same aesthetic experience, I am denying that contextualist considerations play a role in determining the aesthetic properties of an object. This, however, is incorrect. The contextualist does indeed hold that two indistinguishable objects may differ aesthetically. But the contextualist’s point concerns properties that are actually possessed by the objects in question. The fictionalist stance does not therefore clash with the contextualist framework because it does not deny that two perceptually identical artworks may have different aesthetic properties. It simply assumes that it is possible to fictionally engage with an artwork while perceptually engaging with an object perceptually indistinguishable indistinguishable from that artwork.[22]

In fact, the fictionalist stance I described here is not only compatible with contextualism. Rather, it figures in one of the most common ways to introduce and defend such a position. Suppose I am in front of Titian’s Tarquin and Lucretia (1570). In order to make the point that Titian’s reds are only expressive if considered in light of the tradition of Italian painting in which he developed his style, I ask you to imagine the Titian to be an expressionist work. Considered within such a tradition, Titian’s colors seem markedly duller. One way to describe what is going on here is to say that you are make-believe-taking the object in front of you, with all of its perceptual properties, to have a quite different causal history and hence substantially different art-historical properties from the ones you know it has. This is exactly what would need to happen in order for perfect copies to afford an aesthetic surrogate of an original.

I will now move to a second, more serious objection. A problem ensues if I refuse to commit to Currie’s ontological agenda. His claim that a perfect copy of a painting would result in an identical aesthetic experience as the one afforded by the original rests on the instance multiplicity hypothesis, that is, that from a modern edition of Oliver Twist we get the same aesthetic experience as we would get from Dickens’s manuscript, as they both count as instances of the work. In the case of paintings, a perfect copy would count as an instance of the work; hence it could be used as a legitimate aesthetic surrogate. However, if we give up the instance multiplicity hypothesis, and embrace instead the fictionalist proposal, an aspect of the experience of the original will be missing from the experience of the perfect copy. This is the experience of being in front of, or in perceptual contact with, the very object possessing such-and-such aesthetic properties.

For brevity, I will call this the aura experience, with reference to Walter Benjamin’s much-discussed concept of aura. The crucial question is this: Is aura experience an aesthetic aspect of the experience of an original?[23] I do not wish to settle this question here. I will, however, draw the consequences of both a positive and negative answer. If aura experience is part of the aesthetic experience of art, then perfect copies are indeed imperfect aesthetic surrogates of an original. However, and quite importantly, they are not imperfect surrogates for the same reasons as painted copies, that is, they do not corrupt aesthetic understanding. Such corruption derived, it will be remembered, from the lack of a reliable set of non-perceptual properties on which the viewer would bring to bear appreciative categories. But perfect copies, as shown by the Veronese case, do not present this problem. If, on the other hand, aura experience is not part of the aesthetic experience of artworks, then the absence of the thrill of being in front of the real thing does not hinder a perfect copy to function as a perfect aesthetic surrogate of the original. It may, however, be the reason why we still want to see it in person.[24]


Matteo Ravasio,

Matteo Ravasio is a PhD candidate at the University of Auckland, where he works under the supervision of Stephen Davies. His doctoral thesis is devoted to the problem of musical expressiveness, although he also maintains interests in everyday aesthetics, environmental aesthetics, and philosophical issues related to the visual arts.

Published on March 8, 2018.



[1] I will use ‘replica’ and ‘copy’ interchangeably to mean reproductions that do not involve the original artist. However, it is important to note that, in art-historical parlance, a replica is a version of a painting produced by the author of the original whereas a copy is a reproduction made by someone other than the original’s maker. I do not deny there may be interesting observations to be made with regard to replicas specifically. My interest here is to examine the capacity of a reproduction to function as a surrogate of an original, regardless of its author. 

[2] Nathaniel Dunn, Curatorial notes to the exhibition Replicating Genius: Impressionism 1874, (11th to 17th November 2016, Gus Fisher Gallery, Auckland, New Zealand).

[3] Ibid.

[4] While I recognize a distinction between aesthetic and artistic properties, in this paper I use ‘aesthetic’ to refer to both.

[5] Ibid.

[6] The production of copies was and still is considered an important pedagogical tool in the training process of Western artists. While the practice of copying paintings may be partly motivated by its mere capacity to improve a painter’s technique, aesthetic considerations are also crucial in a painter’s choices regarding which works she or he should copy and which aspects of them she or he should focus on. For some examples of this practice, see Theodore Reff, “Degas’s Copies of Older Art,” The Burlington Magazine, 105, 723 (1963), 241-51; Theodore Reff, “New Light on Degas’s Copies,” The Burlington Magazine, 106, 735 (1964), 250-59; Jeffrey M. Muller, “Ruben’s Theory and Practice of the Imitation of Art,” The Art Bulletin, 64, 2, (1982), 229-47; Lisa Pon, Raphael, Dürer and Marcantonio Raimondi. Copying and the Italian Renaissance Print (New Haven and London: Yale University Press, 2004).

[7] For a description of both the aesthetic motivations and technical aspects behind this replica, see Bruno Latour, and Adam Lowe, “The Migration of the Aura or How to Explore the Original Through its Fac Similes,” in Switching Codes, ed. Thomas Bartscherer (Chicago: University of Chicago Press, 2010).

[8] See Robert Hopkins, “Reproductive Prints as Aesthetic Surrogates,” The Journal of Aesthetics and Art Criticism, 73, 1 (2015), 12-21.

[9] Sherri Irvin, “Forgery and the Corruption of Aesthetic Understanding,” Canadian Journal of Philosophy, 37, 2 (2007), 283-304.

[10] Ibid., p. 301.

[11] Hope B. Werness, “Han Van Meegeren fecit,” in The Forger’s Art, ed. Denis Dutton (Berkeley and Los Angeles: University of California Press, 1983), pp. 54-56.

Lawrence Gowing, author of one of the most important critical studies on Vermeer, had already written in 1952: “After the Second World War it was found that a catastrophe had overtaken criticism. Half a dozen fabrications had been accepted with enthusiasm by leading authorities. Yet the confusion itself perhaps shed light on accepted ides about the painter.” Lawrence Gowing, Lawrence, Vermeer (London: Giles de la Mare, 1997); ref on p.66.

[12] Jerrold Levinson distinguishes between referential and inventive forgeries. See his “Autographic and Allographic Art Revisited,” Philosophical Studies, 38, 4 (1980), 367-383. Referential forgeries are copies of existing artworks presented as the original. Inventive forgeries, such as Van Meegeren’s Vermeers, are new works with an intentionally mistaken attribution and succeed insofar as they mimic the author’s or period’s style. Given this distinction, it might be asked whether more could be said regarding the different ways in which a known forgery functions as a perceptual benchmark in the two cases. That it may do so in the case of inventive forgeries is testified by the beneficial effect on criticism that followed the discovery of Van Meegeren’s fraud. To explain how it may do so would require further work, as my framework is more readily applicable to the case of referential forgeries.

[13] Fabritius had worked in Rembrandt’s workshop.

[14] A note of caution is in order here. It would be an overstatement to claim that the harm of the museum of copies is just like the harm caused by a museum of forgeries. Forgeries not only corrupt aesthetic understanding, they also undermine the trust we have in the experts who we thought able to distinguish an original from a fake. In this sense, the harm they may cause is far greater than the one caused by copies.

[15] Hopkins, “Reproductive Prints as Aesthetic Surrogates,” p.14.

[16] Barbara Savedoff, “Looking at Art through Photographs,” The Journal of Aesthetics and Art Criticism, 51, 3 (1993), 455-462.

[17] Ibid., p. 460.

[18] Hopkins, “Reproductive Prints as Aesthetic Surrogates,” p.15.

[19] This difference between the properties of the copy and the ones of the original is different from the one highlighted by Hopkins concerning transparency. Whereas the latter may include perceptual properties, for example, Mount Cook is majestic while its photograph is not, the former is limited to non-perceptual properties such as ones related to art-historical significance.

[20] Gregory Currie, “The Authentic and the Aesthetic,” American Philosophical Quarterly, 22, 2 (1985), 153-160; Gregory Currie, An Ontology of Art (London: Macmillan, 1989). Another author who embraces something akin to the instance multiplicity hypothesis in the case of painting is Eddy Zemach. See his “No Identification without Evaluation,” British Journal of Aesthetics, 26, 3 (1986), 239-251. For criticism, see Jerrold Levinson “Zemach on Paintings,” British Journal of Aesthetics, 27, 3 (1987), 278-283.

[21] Currie, “The Authentic and the Aesthetic,” p.157.

[22] The point that two perceptually indistinguishable objects may nonetheless differ aesthetically is made, against Goodman, in Robert Hopkins, “Aesthetics, Experience, and Discrimination,” The Journal of Aesthetics and Art Criticism, 63, 2 (2005), 119-133.

[23] Currie answers to this question in the negative in An Ontology of Art, p. 102.

[24] This paper has benefited from many insightful comments and helpful suggestions. My gratitude goes to an anonymous reviewer, Stephen Davies, Julian Dodd, Nathaniel Dunn, Ivan Gaskell, Karen Gover, Fred Kroon, Jerrold Levinson, and various audience members at the American Society of Aesthetics 2017 Annual Meeting in New Orleans.